System for customizing in-game character animations by players
Abstract
System and methods for using a deep learning framework to customize animation of an in-game character of a video game. The system can be preconfigured with animation rule sets corresponding to various animations. Each animation can be comprised of a series of distinct poses that collectively form the particular animation. The system can provide an animation-editing interface that enables a user of the video game to make modifications to at least one pose or frame of the animation. The system can realistically extrapolate these modifications across some or all portions of the animation. In addition or alternatively, the system can realistically extrapolate the modifications across other types of animations.
Claims
exact text as granted — not AI-modifiedWhat is claimed is:
1. A computer-implemented method for customizing animation of an in-game character of a video game, the method comprising:
as implemented by one or more hardware processors of a computing system configured with specific computer-executable instructions,
executing a game application on a client computing device, wherein the game application comprises game data, wherein the game data comprises an animation rule set including a series of character poses defining animation of a first action of an in-game character within the game application, wherein an in-game character comprises a plurality of rigid bodies connected by a plurality of joints, and wherein each character pose of the series of character poses is a particular arrangement of the rigid bodies and joints of the in-game character; and
during runtime of the game application:
identifying one or more character poses associated with animation of the first action;
receiving, from a user of the game application, a modification to a first character pose of the one or more character poses, wherein the modification comprises a change to at least one position, angle, or velocity associated with the arrangement of the rigid bodies and joints of the in-game character;
generating one or more modified character poses for the first action based at least in part on the modification to the first character pose, wherein the one or more modified character poses for the first action are generated using a motion generation machine learning model; and
updating the animation rule set to generate a modified animation rule set for the first action, wherein the one or more modified character poses for the first action replace the corresponding one or more unmodified character poses for the first action, wherein the modified animation rule set defining the modified one or more character poses generated by the character within a virtual environment of the game application when performing the first action.
2. The computer-implemented method of claim 1 , wherein the animation rule set is a first animation rule set, wherein the series of character poses is a first series of character poses, wherein the game data further comprises a second animation rule set including a second series of character poses defining animation of a second action, wherein the method further comprises:
generating one or more modified character poses for the second action based at least in part on the modification to the first character pose, wherein the one or more modified character poses for the second action are generated using a motion generation machine learning model; and
updating a second animation rule set to generate a modified second animation rule set for the second action, wherein the one or more modified character poses for the second action replace corresponding one or more unmodified character poses for the second action.
3. The computer-implemented method of claim 2 , wherein the first action is a first type of animation and the second action is a second type of animation that is different from the first type of animation.
4. The computer-implemented method of claim 3 , wherein the first type of animation is one of a locomotion or an asynchronous motion, and wherein the second type of animation is the other one of the locomotion or the asynchronous motion.
5. The computer-implemented method of claim 4 , wherein the locomotion comprises at least one of running, jumping, hopping, crawling, marching, climbing, galloping, sliding, leaping, hopping, or skipping.
6. The computer-implemented method of claim 4 , wherein the asynchronous motion comprises at least one of standing up, sitting down, punching, kicking, swing a sword, or bending.
7. The computer-implemented method of claim 3 , wherein the first type of animation is one of a first type of locomotion, and wherein the second type of animation is a second type of locomotion.
8. The computer-implemented method of claim 1 , further comprising:
post-processing the one or more modified character poses for the first action to resolve motion artifacts of the one or more modified character poses for the first action associated with the modification.
9. The computer-implemented method of claim 1 , said identifying the one or more character poses associated with animation of the first action comprises receiving a selection from a user of the game application, wherein the selection indicates the one or more character poses associated with animation of the first action.
10. The computer-implemented method of claim 1 , further comprising:
monitoring gameplay of the game application during the runtime of the game application;
identifying, during the monitoring, an occurrence of a first triggering game state during the runtime of the game application, the identifying being based on defined game conditions within the game application; and
responsive to said identifying the occurrence of the first triggering game state, applying the modified animation rule set during the runtime of the game application causing the in-game character to perform the first action.
11. The computer-implemented method of claim 1 , further comprising:
providing an animation editor interface that provides an interface for the user to select the modification to the to the first character pose of the one or more character poses.
12. The computer-implemented method of claim 11 , wherein the animation editor interface displays a plurality of possible modifications, wherein the modification comprises a set of the plurality of possible modifications.
13. A system comprising:
an electronic data store configured to store computer-executable instructions associated with generating frames of an animation; and
a hardware processor in communication with the electronic data store and configured to execute the computer-executable instructions to at least:
execute a game application on a client computing device, wherein the game application comprises game data, wherein the game data comprises an animation rule set including a series of character poses defining animation of a first action of an in-game character within the game application, wherein an in-game character comprises a plurality of rigid bodies connected by a plurality of joints, and wherein each character pose of the series of character poses is a particular arrangement of the rigid bodies and joints of the in-game character; and
during runtime of the game application:
identify one or more character poses associated with animation of the first action;
receive, from a user of the game application, a modification to a first character pose of the one or more character poses, wherein the modification comprises a change to at least one position, angle, or velocity associated with the arrangement of the rigid bodies and joints of the in-game character;
generate one or more modified character poses for the first action based at least in part on the modification to the first character pose, wherein the one or more modified character poses for the first action are generated using a motion generation machine learning model; and
update the animation rule set to generate a modified animation rule set for the first action, wherein the one or more modified character poses for the first action replace the corresponding one or more unmodified character poses for the first action, wherein the modified animation rule set defining the modified one or more of character poses generated by the character within a virtual environment of the game application when performing the first action.
14. The system claim 13 , wherein the animation rule set is a first animation rule set, wherein the series of character poses is a first series of character poses, wherein the game data further comprises a second animation rule set including a second series of character poses defining animation of a second action, wherein the one or more processors are further configured to:
generate one or more modified character poses for the second action based at least in part on the modification to the first character pose, wherein the one or more modified character poses for the second action are generated using a motion generation machine learning model; and
update a second animation rule set to generate a modified second animation rule set for the second action, wherein the one or more modified character poses for the second action replace corresponding one or more unmodified character poses for the second action.
15. The system claim 14 , wherein the first action is a first type of animation and the second action is a second type of animation that is different from the first type of animation.
16. The system claim 15 , wherein the first type of animation is one of a locomotion or an asynchronous motion, and wherein the second type of animation is the other one of the locomotion or the asynchronous motion.
17. The system claim 16 , wherein the locomotion comprises at least one of running, jumping, hopping, crawling, marching, climbing, galloping, sliding, leaping, hopping, or skipping.
18. The system claim 16 , wherein the asynchronous motion comprises at least one of standing up, sitting down, punching, kicking, swing a sword, or bending.
19. The system claim 15 , wherein the first type of animation is one of a first type of locomotion, and wherein the second type of animation is a second type of locomotion.
20. A non-transitory computer-readable storage medium storing computer-executable instructions that, when executed by one or more computing devices, configure the one or more computing devices to perform operations comprising:
executing a game application on a client computing device, wherein the game application comprises game data, wherein the game data comprises an animation rule set including a series of character poses defining animation of a first action of an in-game character within the game application, wherein an in-game character comprises a plurality of rigid bodies connected by a plurality of joints, and wherein each character pose of the series of character poses is a particular arrangement of the rigid bodies and joints of the in-game character; and
during runtime of the game application:
identifying one or more character poses associated with animation of the first action;
receiving, from a user of the game application, a modification to a first character pose of the one or more character poses, wherein the modification comprises a change to at least one position, angle, or velocity associated with the arrangement of the rigid bodies and joints of the in-game character;
generating one or more modified character poses for the first action based at least in part on the modification to the first character pose, wherein the one or more modified character poses for the first action are generated using a motion generation machine learning model; and
updating the animation rule set to generate a modified animation rule set for the first action, wherein the one or more modified character poses for the first action replace the corresponding one or more unmodified character poses for the first action, wherein the modified animation rule set defining the modified one or more of character poses generated by the character within a virtual environment of the game application when performing the first action.Cited by (0)
No later patents cite this yet.
References (0)
No backward citations on record.